How to measure Performance
Arrival rates and distributions of service requests, processing times, queue sizes and latency (the rate at which requests are serviced)
simulate by building a stochastic queueing model of the system based upon anticipated workload scenarios